Star Ratings rule changes
CMS has proposed a large change to how Medicare Advantage Star Ratings are calculated: a much smaller measure set, no reward factor, and less protection from the improvement measures. Below: what the rule changes, and our estimate of what it would do to today's ratings.
Contract Year 2028 Policy and Technical Changes (proposed rule CMS-4214-P2) · Proposed
Filed for public inspection November 20, 2026 · scheduled for publication in the Federal Register November 25, 2026 · comments due January 19, 2027 (60 days after filing) · FR Doc. 2026-23960
Timing: the measure and method changes would apply to the 2028 measurement year, the 2030 Star Ratings (published October 2029), and the quality bonus payments for 2031. Comments are due 60 days after the November 20, 2026 public inspection filing. CMS may finalize all, some or none of the measure removals.
The rule (PDF) · CMS fact sheet · earlier: Contract Year 2027 and Certain Contract Year 2026 Policy and Technical Changes (final rule) (91 FR 17384)
i
For each contract that would fall below 4 stars: 5% of its risk-adjusted Part A and B payments for a year (base payment per member per month × average risk score × member months, from CMS's 2024 plan payment data). The real bonus raises the county benchmark by 5% (10% in some counties), and only part of it reaches the plan's payment, so this is a rough size, not a forecast. 4 of the contracts have no payment data and aren't counted. Not counted either: the rebate share also falls with the rating (70% of savings at 4.5 stars or more, 65% at 3.5 or 4, 50% at 3 or less).
CMS's own estimate in the rule, which includes those effects and growth: −$11.5 billion in 2031 rising to −$18.9 billion in 2037, −$107 billion over 2028–2037 (about 1.2% of Medicare's payments to private plans).
Our estimate: today's (2027) measure stars, recomputed under the proposed rules. It isn't a forecast of the 2030 ratings, which will also reflect three more years of performance and cut-point changes.
What the rule changes
| Change | First Star Ratings | What it does | Why it matters |
|---|---|---|---|
| Nineteen more measures removedmeasures | 2030 | Sixteen process measures and three patient experience measures leave with the 2030 Star Ratings (2028 measurement year): breast and colorectal cancer screening, flu vaccine, monitoring physical activity, both care for older adults measures, osteoporosis management, diabetes eye exam, kidney health evaluation, reducing the risk of falling, improving bladder control, transitions of care, depression screening, getting appointments and care quickly, rating of health plan, rating of drug plan, statin use in diabetes, opioids with benzodiazepines, and anticholinergic polypharmacy. | The rating would rest on 15 measures: eight outcome or intermediate outcome measures, three patient experience measures, two process measures and the two improvement measures. The improvement measures (weight 5 each) become about a quarter of the overall rating's weight. |
| Reward factor removedmethod | 2030 | The 0.1 to 0.4 star reward for high, consistent measure stars ends with the 2030 Star Ratings. | Every contract that earns a reward today loses it; CMS says this puts downward pressure on ratings. |
| Improvement hold harmless only at 4.5 stars and upmethod | 2030 | Today a contract rated 4 stars or more without the improvement measures can't be pulled down by them. From the 2030 Star Ratings that protection would apply only at 4.5 stars or more. | 4-star contracts whose improvement measures are weak could fall to 3.5 stars and lose the quality bonus. |
| Blood pressure measure updatedmeasures | 2031 | Controlling Blood Pressure is replaced by NCQA's updated version (BPC-E), which also counts members with one hypertension visit and a blood pressure prescription, and moves to electronic clinical data reporting. It counts with weight 1 in its first year (2031 Star Ratings), then 3. | A wider denominator can move contracts' scores and cut points for a year. |
| Cut points adjusted for the ADVaNCE MA modelmethod | 2031 | Contracts in the proposed ADVaNCE MA model (randomly chosen states from 2029) would get no Star Ratings. For everyone else's cut points, CMS would use their scores adjusted for the model's estimated effect, and leave them out of the improvement measures, reward factor and CAI. | Keeps the model from shifting other contracts' cut points. |
| ADVaNCE MA Diamond Ratingsmodel | 2031 | In the model's states, a Diamond Rating would replace Star Ratings and the quality bonus: performance compared within each state, by parent organization, in separate general enrollment and D-SNP cohorts, shown in finer steps than half stars (tenths, for example) and paid on a sliding scale rather than at a single 4-star line. The states would be picked at random from seven groups of similar states and named in the final rule. | Parents with contracts in the chosen states would be rated against local competitors, not nationally. |
| Empty domains not showndisplay | 2030 | Medicare Plan Finder would show only domains that still have measures. CMS says it may redraw the domains later. | Some of today's domains could disappear from Plan Finder. |
| Preliminary ratings stay privateother | 2030 | Plans and their contractors could not publish preliminary Star Ratings data from the plan preview periods before CMS releases the ratings in October. | — |
| Disaster policy clarifiedother | 2030 | When a disaster-affected contract has too few measures for a rating calculated one of the two ways (with or without new measures), it gets the rating that can be calculated. | — |
Already final
The proposal builds on the 2027 rule CMS finalized in April 2026:
- Eleven measures removed (2028 Star Ratings): Three measures leave with the 2028 Star Ratings (statin therapy for cardiovascular disease and both call center interpreter measures) and eight more with the 2029 Star Ratings: special needs plan care management, customer service, rating of health care quality, complaints about the plan, members choosing to leave the plan (each counted once though rated in Part C and D), both appeals measures and Medicare Plan Finder price accuracy.
- Depression screening added (2029 Star Ratings): Depression Screening and Follow-up joins the Star Ratings with the 2029 release, as a process measure (the average of the screening and follow-up rates).
- Reward factor kept (2027 Star Ratings): The reward factor stays in place of the planned health equity index reward.
The measure seti
How the weights are countedWeights are CMS's 2027 weights: outcome and intermediate outcome measures 3, patient experience measures 2, process measures 1, and each improvement measure 5. The share counts every measure in an MA-PD contract's overall rating once (the Part D complaints and members-leaving measures repeat Part C's and count once). Measures not in the 2027 ratings (the medication therapy management review rate, depression screening) have no 2027 weight. The rule doesn't propose new weights; if CMS changes them, the shares change.
i
Weights are CMS's 2027 weights: outcome and intermediate outcome measures 3, patient experience measures 2, process measures 1, and each improvement measure 5. The share counts every measure in an MA-PD contract's overall rating once (the Part D complaints and members-leaving measures repeat Part C's and count once). Measures not in the 2027 ratings (the medication therapy management review rate, depression screening) have no 2027 weight. The rule doesn't propose new weights; if CMS changes them, the shares change.
What's left leans on outcomes and medication adherence. Of the 2030 weight in an MA-PD's overall rating (at 2027 weights, without the measures not yet rated), the three Part D adherence measures carry 22%, the two Health Outcomes Survey measures 15%, blood sugar, blood pressure and readmissions 22%, and the three patient experience surveys 15%. Cancer screenings, care for older adults, and most of the HEDIS process measures plans work on today would no longer count.
| Measure | Part | Type | 2027 weight | Status | Note |
|---|---|---|---|---|---|
| C04 Improving or Maintaining Physical Health | C | Outcome | 3 | Kept | |
| C05 Improving or Maintaining Mental Health | C | Outcome | 3 | Kept | |
| C12 Diabetes Care – Blood Sugar Controlled | C | Intermediate outcome | 3 | Kept | |
| C14 Controlling High Blood Pressure | C | Intermediate outcome | 3 | Kept | Replaced by NCQA's BPC-E (wider denominator, electronic reporting); weight 1 in 2031, then 3 (2031, proposed). |
| C17 Plan All-Cause Readmissions | C | Outcome | 3 | Kept | |
| C20 Follow-up after Emergency Department Visit for People with Multiple High-Risk Chronic Conditions | C | Process | 1 | Kept | |
| C21 Getting Needed Care | C | Patients' experience of care | 2 | Kept | |
| C26 Care Coordination | C | Patients' experience of care | 2 | Kept | |
| C29 Health Plan Quality Improvement | C | Improvement | 5 | Kept | |
| D04 Drug Plan Quality Improvement | D | Improvement | 5 | Kept | |
| D06 Getting Needed Prescription Drugs | D | Patients' experience of care | 2 | Kept | |
| D08 Medication Adherence for Diabetes Medications | D | Intermediate outcome | 3 | Kept | |
| D09 Medication Adherence for Hypertension (RAS antagonists) | D | Intermediate outcome | 3 | Kept | |
| D10 Medication Adherence for Cholesterol (Statins) | D | Intermediate outcome | 3 | Kept | |
| Medication Therapy Management Program Completion Rate for Comprehensive Medication Reviewnot in the 2027 ratings | D | Process | — | Kept | Not in the 2027 Star Ratings; CMS reports a 44% average MA-PD completion rate for 2025 |
| C01 Breast Cancer Screening | C | Process | 1 | Proposed: removed from 2030 | |
| C02 Colorectal Cancer Screening | C | Process | 1 | Proposed: removed from 2030 | |
| C03 Annual Flu Vaccine | C | Process | 1 | Proposed: removed from 2030 | |
| C06 Monitoring Physical Activity | C | Process | 1 | Proposed: removed from 2030 | |
| C08 Care for Older Adults – Medication Review | C | Process | 1 | Proposed: removed from 2030 | |
| C09 Care for Older Adults - Functional Status Assessment | C | Process | 1 | Proposed: removed from 2030 | |
| C10 Osteoporosis Management in Women who had a Fracture | C | Process | 1 | Proposed: removed from 2030 | |
| C11 Diabetes Care – Eye Exam | C | Process | 1 | Proposed: removed from 2030 | |
| C13 Kidney Health Evaluation for Patients with Diabetes | C | Process | 1 | Proposed: removed from 2030 | |
| C15 Reducing the Risk of Falling | C | Process | 1 | Proposed: removed from 2030 | |
| C16 Improving Bladder Control | C | Process | 1 | Proposed: removed from 2030 | |
| C19 Transitions of Care | C | Process | 1 | Proposed: removed from 2030 | |
| C22 Getting Appointments and Care Quickly | C | Patients' experience of care | 2 | Proposed: removed from 2030 | |
| C25 Rating of Health Plan | C | Patients' experience of care | 2 | Proposed: removed from 2030 | |
| D05 Rating of Drug Plan | D | Patients' experience of care | 2 | Proposed: removed from 2030 | |
| D11 Statin Use in Persons with Diabetes (SUPD) | D | Process | 1 | Proposed: removed from 2030 | |
| D12 Concurrent Use of Opioids and Benzodiazepines (COB) | D | Process | 1 | Proposed: removed from 2030 | |
| D13 Polypharmacy: Use of Multiple Anticholinergic Medications in Older Adults (Poly-ACH) | D | Process | 1 | Proposed: removed from 2030 | |
| Depression Screening and Follow-upnot in the 2027 ratings | C | Process | — | Proposed: removed from 2030 | Added from 2029 by the 2027 rule. |
| C18 Statin Therapy for Patients with Cardiovascular Disease | C | HEDIS | 1 | Removed from 2028 (final) | |
| C32 Call Center – Foreign Language Interpreter and TTY Availability | C | Call Center Monitoring | 2 | Removed from 2028 (final) | |
| D01 Call Center – Foreign Language Interpreter and TTY Availability | D | Call Center Monitoring | 2 | Removed from 2028 (final) | |
| C07 Special Needs Plan (SNP) Care Management | C | Plan Reporting Requirements | — | Removed from 2029 (final) | |
| C23 Customer Service | C | CAHPS | 2 | Removed from 2029 (final) | |
| C24 Rating of Health Care Quality | C | CAHPS | 2 | Removed from 2029 (final) | |
| C27 Complaints about the Health Plan | C | Complaints Tracking Module | 2 | Removed from 2029 (final) | |
| C28 Members Choosing to Leave the Plan | C | Enrollment Data | 2 | Removed from 2029 (final) | |
| C30 Plan Makes Timely Decisions about Appeals | C | Independent Review Entity | 2 | Removed from 2029 (final) | |
| C31 Reviewing Appeals Decisions | C | Independent Review Entity | 2 | Removed from 2029 (final) | |
| D02 Complaints about the Drug Plan | D | Complaints Tracking Module | 2 | Removed from 2029 (final) | |
| D03 Members Choosing to Leave the Plan | D | Enrollment Data | 2 | Removed from 2029 (final) | |
| D07 MPF Price Accuracy | D | Prescription Drug Events | 1 | Removed from 2029 (final) |
What it would do to today's ratingsi
How we estimate it
Method: each contract's 2027 rating is recomputed with CMS's own arithmetic (weighted mean of measure stars, reward factor, categorical adjustment index, improvement-measure rule and rounding), the same calculation that reproduces CMS's published 2027 overall ratings for 495 of 508 contracts. Then the rule's changes are applied one at a time, each step on top of the one before, so the table shows what each part does. This is the approach CMS used for its own estimate in the rule (2027 data with the measure set updated).
Held at today's values, because public data can't recompute them: measure stars and cut points (removing other measures doesn't change them); the two improvement measures' stars, which CMS would recalculate over the smaller measure set; and the categorical adjustment index values, which CMS recalculates every year. When the measure set changes, the reward-factor thresholds move with the market (they're percentiles of all contracts' scores). Measures not in the 2027 ratings can't be counted.
Rating: the overall rating for contracts that have one, otherwise the Part C summary (what decides the quality bonus). Part D-only plans aren't shown. Where our rebuild of today's rating differs from CMS's published one (13 contracts), the table marks it; the steps start from the rebuilt rating so each step's effect is the rules' alone.
Order: each step adds to the ones before, so how much of the change a step shows depends on the order; the full proposal's result doesn't.
Small contracts: with 15 measures, 19 small contracts (131,160 members in all) would have scores on fewer than half of them and might get no rating at all; they're shown anyway.
Check against CMS: CMS's estimate in the rule, on the same 2027 data: 36% of MA-PD contracts keep their overall rating, 6% rise and 58% fall; 2% of MA contracts gain the quality bonus and 20% lose it. Ours, for the 508 contracts with an overall rating: 36% same, 3% up, 61% down; 1% gain the bonus and 21.7% lose it. CMS recomputes the improvement measures and CAI, which we can't; that likely explains why a few more contracts rise in its estimate.
i
Method: each contract's 2027 rating is recomputed with CMS's own arithmetic (weighted mean of measure stars, reward factor, categorical adjustment index, improvement-measure rule and rounding), the same calculation that reproduces CMS's published 2027 overall ratings for 495 of 508 contracts. Then the rule's changes are applied one at a time, each step on top of the one before, so the table shows what each part does. This is the approach CMS used for its own estimate in the rule (2027 data with the measure set updated).
Held at today's values, because public data can't recompute them: measure stars and cut points (removing other measures doesn't change them); the two improvement measures' stars, which CMS would recalculate over the smaller measure set; and the categorical adjustment index values, which CMS recalculates every year. When the measure set changes, the reward-factor thresholds move with the market (they're percentiles of all contracts' scores). Measures not in the 2027 ratings can't be counted.
Rating: the overall rating for contracts that have one, otherwise the Part C summary (what decides the quality bonus). Part D-only plans aren't shown. Where our rebuild of today's rating differs from CMS's published one (13 contracts), the table marks it; the steps start from the rebuilt rating so each step's effect is the rules' alone.
Order: each step adds to the ones before, so how much of the change a step shows depends on the order; the full proposal's result doesn't.
Small contracts: with 15 measures, 19 small contracts (131,160 members in all) would have scores on fewer than half of them and might get no rating at all; they're shown anyway.
Check against CMS: CMS's estimate in the rule, on the same 2027 data: 36% of MA-PD contracts keep their overall rating, 6% rise and 58% fall; 2% of MA contracts gain the quality bonus and 20% lose it. Ours, for the 508 contracts with an overall rating: 36% same, 3% up, 61% down; 1% gain the bonus and 21.7% lose it. CMS recomputes the improvement measures and CAI, which we can't; that likely explains why a few more contracts rise in its estimate.
Most of the drop comes from two places. The measures the 2027 rule already removed: contracts score well on complaints, appeals, call centers and members leaving (3.7 to 4.4 stars on average, against 3.4 across all measures), so taking them out lowers about 9 in 10 contracts' average. That part is already final and applies from the 2029 Star Ratings whether or not this proposal goes ahead: with today's measure stars, members in 4-star-or-better contracts would fall by 7.3M on it alone. CMS's estimate for that rule was close to neutral, but it measured a package that also kept the reward factor instead of the planned health equity reward, on older data. The end of the reward factor is the second. The hold harmless change then moves 4-star contracts whose improvement measures are weak down to 3.5.
| Step (each includes those above) | Lower than today | Higher | Contracts at 4★+ | Members at 4★+ | Member-weighted average |
|---|---|---|---|---|---|
| Today's rating (rebuilt) | — | — | 196 of 515 | 25.6M | 3.99 |
| Measures already removed by final rules | 226 | 12 | 144 of 515 | 18.3M | 3.72 |
| Plus the proposed measure removals | 250 | 30 | 156 of 515 | 18.5M | 3.68 |
| Plus no reward factor | 301 | 17 | 118 of 515 | 11.6M | 3.52 |
| Plus hold harmless only at 4.5 stars (the full proposal) | 314 | 17 | 91 of 515 | 9.7M | 3.49 |
Ratings today and under the proposal
| Rating | Contracts today | Proposal | Members today | Proposal |
|---|---|---|---|---|
| 5 | 13 | 7 | 0.2M | 0.1M |
| 4.5 | 67 | 12 | 12M | 0.2M |
| 4 | 116 | 72 | 13.4M | 9.3M |
| 3.5 | 161 | 162 | 6.2M | 17.1M |
| 3 | 125 | 165 | 3.1M | 6.3M |
| 2.5 | 32 | 82 | 0.3M | 2M |
| 2 | 1 | 14 | 0M | 0.1M |
| 1.5 | 0 | 1 | 0M | 0M |
What it means
- The bonus line gets harder to reach. With today's performance, the share of members in 4-star contracts would fall from 73% to 28%. Plans keeping the bonus would need higher measure stars than today.
- Improvement counts for more. The two improvement measures become about a quarter of the rating, and the hold harmless would no longer protect 4-star contracts. A year of slipping on the remaining measures costs more than it does today.
- Fewer places to earn stars. Screening and other process-measure gap closure would no longer move Star Ratings; medication adherence, blood sugar and blood pressure control, readmissions, the Health Outcomes Survey and three patient experience surveys would.
- It's a proposal. CMS says it may finalize all, some or none of the removals. It finalized last year's rule in April.
By size of parent organizationi
How parents are groupedParent organizations ranked by members in their rated contracts (latest month), then grouped. Each group adds up its contracts' figures; the average rating is weighted by members. Bonus: the rough size described above.
i
Parent organizations ranked by members in their rated contracts (latest month), then grouped. Each group adds up its contracts' figures; the average rating is weighted by members. Bonus: the rough size described above.
Where the change lands, without naming organizations.
| Parents | Members | Average rating today | Proposal | Members at 4★+ today | Proposal | Contracts losing the bonus | Rough bonus lost a year |
|---|---|---|---|---|---|---|---|
| Top 5 parents 5 parents | 24.3M | 4.06 | 3.54 | 19.5M 80% | 7.7M 32% | 38 | $6.9B |
| Parents 6–25 20 parents | 7.3M | 3.87 | 3.37 | 4.3M 59% | 1.2M 17% | 41 | $1.4B |
| Parents 26–75 50 parents | 3M | 3.74 | 3.42 | 1.6M 54% | 0.6M 20% | 25 | $524M |
| Parents 76 and below 57 parents | 0.5M | 3.59 | 3.31 | 0.2M 36% | 0.1M 25% | 7 | $23.6M |
By organization and contract: the estimate for every parent organization and contract, each step of the proposal, and the rough bonus at stake, is for members. Sign in to see every organization and contract.